--- On Tue, 1/12/10, Imago <[email protected]> wrote:
...
> But in my opinion even fragile filtering is better then
> none at all. Because 
> while some could get in the population en masse wouldn't be
> able to.

This statement and the responses to it illustrate the reason I changed the 
subject line to a generic one. Sadly you are not the only person who thinks 
this way and it is totally false! 

I hasten to add that this tirade is not specifically directed at you, but at 
the general notion that you express here. You propose a very bad idea which is 
far too highly respected by the naive and inexperienced. 

This approach is far worse than "security by obscurity", rather it is merely 
"feelgood security". By analogy it is like installing dummy fire extinguishers 
rather than ones which work because "at least" the dummy ones make everyone 
feel more secure. I suggest you ask your local fire department about the wisdom 
of dummy fire extinguishers!

The problem is that no security measure comes without a penalty. That penalty 
is usually paid by your desired (and honest) users. The penalty is usually not 
apparent before deployment, but the penalty is very real and often deadly.

Typically what suffers is the robustness of the system (frequent crashes, loss 
of data, etc.) or simple lack of usefulness for its intended job. In this case 
there may well be very good reasons for people to use other viewers on 
occasion. 

I count 6 different Linux based viewers that I use regularly and a similar (but 
different) set that I use with Windows. Each of these have their strengths and 
weaknesses, and at one time or another each was the ONLY practical way of 
performing some function that I needed to do!

There is a time for bluffing, but planning your security system is not one of 
them. Especially if those being "bluffed" are on your team!

Probably your best guide is the OSGrid grid network which is intended as a 
"test bed" for the OpenSim software. They use the normal username/password 
system with a fairly simple registration system for access and apparently have 
few problems. I am sure that a direct email to one of the administrators of 
that network would bring some very good advice based on real world experience.

My suspicion is that an investment in regular and complete database backups 
(ask me how I know that!!!) would be far more useful and cost less effort 
overall than restricting user access based on some hypothetical problem with 
unknown viewer software.

Karen


      
_______________________________________________
Opensim-users mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/opensim-users

Reply via email to